packetforwarding

Packet forwarding is the process by which a network device moves an IP packet from its ingress interface toward its destination. It is distinct from routing, which determines the path by selecting routes based on routing protocols and policies; forwarding is the act of sending the packet along the chosen path using a forwarding table.

Forwarding decisions occur in the forwarding plane of a device, typically a router or a layer 3 switch. The device uses the destination address in the packet header to perform a lookup in a routing or forwarding table to determine the next hop and the egress interface. For IP traffic, the lookup often uses the longest-prefix match. In MPLS networks, labels may be used to steer the packet instead of IP addresses.

Once the next hop and exit interface are determined, the device may perform address resolution to obtain the link-layer address, decrement the packet’s TTL, and transmit the packet on the selected interface. Some devices use store-and-forward forwarding, buffering the entire packet before forwarding, while others employ cut-through or fast-forward methods to begin sending before the full packet is received. Forwarding paths are subject to security and policy controls such as access control lists and Network Address Translation, which may modify or block packets.

Performance and scalability considerations include the speed of lookups, memory bandwidth, and hardware support such as TCAMs in switches or specialized forwarding engines. Modern devices separate control logic (routing protocols and topology computation) from the forwarding plane to improve efficiency and scalability. Forwarding behavior is influenced by quality of service, congestion management, and buffering, which affect overall network performance in high-traffic environments.